The goals of this research program are to advance our understanding of the language impairments in autism, and how these impairments are related to social-cognitive deficits, specifically in their of mind, that are at the core of the syndrome. The specific focus of this investigation is on autistic individuals' knowledge of the linguistic properties of cognition verbs (e.g., think, know). The primary hypothesis is that autism involves specific deficits in knowledge about the syntactic, semantic, and pragmatic aspects of cognition verbs, and that these deficits are closely linked to their deficits in theory of mind. This hypothesis will be tested in a series of 14 experiments in which autistic subjects' performance will be compared to control groups of mentally retarded/learning disabled and normal subjects matched on standardized tests of vocabulary comprehension (PPVT-R), syntactic comprehension (CELF-R Sentence Structure) expressive language (CELF-R Formulated Sentences) and digit span. The experiments will test the subjects knowledge of syntactic (Experiments 1-5), semantic (Experiments 6-8), pragmatic (Experiments 9-10), and prosodic (Experiments 11-13) characteristics of cognition verbs. Within each experiment linguistic knowledge of cognition verbs will be explicitly compared to control items (e.g., other verbs that take sentence complements) to test whether autism involves impairments specific to cognition verbs or to the linguistic characteristics in general. In addition, subjects in each experiment will be given a theory of mind task battery. The relationship between performance on the linguistic task and the theory of mind tasks will be explored using correlational and non-parametric approaches in order to examine more closely the direction of the relationship between these domains. The final experiment (Experiment 14) will investigate the relationships among knowledge of syntactic, semantic and pragmatic characteristics of cognition verbs. Taken together, these experiments will advance our understanding of the language impairments in autism within a particular lexical domain. Furthermore, these experiments will explore whether autistic individuals use their knowledge about the structure of language, particularly embedded sentence complements to promote their understanding of representational mental states.